For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).
Mojo struct
CausalConv1DVarlenFwdRows
struct CausalConv1DVarlenFwdRows[activation: StaticString, channels_last: Bool = False, use_residual: Bool = False]
causal_conv1d_varlen_fwd in a mixed verify graph.
The verify width K is the length of verify_width, whose contents
are never read. A row longer than K + 1 tokens is a prefill row: it
writes its state window ending at its last real token whatever
update_state says; the flag keeps its meaning for shorter rows. A
launch with no prefill row runs main's convolution.
Tensor Shapes:
As causal_conv1d_varlen_fwd, plus
- verify_width : [K] int64
Parameters
- activation (
StaticString): Activation function - "none" or "silu". - channels_last (
Bool): If True,xandoutputare tokens-major (total_seqlen, dim) instead of (dim, total_seqlen). - use_residual (
Bool): If True, addsx[d, s]to the convolution sum at each output position, beforeactivation.
Implemented traits
Methods
execute
static def execute(output: T, x: T, weight: T, bias: T, conv_states: T, query_start_loc: T, cache_indices: T, has_initial_state: T, update_state: T, verify_width: T, ctx: DeviceContext)